Modeling navigation patterns of visitors of unstructured websites

Abstract

In this paper we describe a practical approach for modeling navigation patterns of visitors of unstructured websites. These patterns are derived from web logs that are enriched with 3 sorts of information: (1) content type of visited pages, (2) visitor type, and (3) location of the visitor. We developed an intelligent Text Mining system, iTM, which supports the process of classifying web pages into a number of pre-defined categories. With help of this system we were able to reduce the labeling effort by a factor 10-20 without affecting the accuracy of the final result too much. Another feature of our approach is the use of a new technique for modeling navigation patterns: navigation trees. They provide a very informative graphical representation of most frequent sequences of categories of visited pages. © 2006 Springer-Verlag London.